Pennsylvania Patent of the Month – June 2023

Neuraville LLC, a leading innovator in the field of artificial intelligence (AI) and neural networks, has developed a groundbreaking method for automatically generating neural network architectures. Their cutting-edge research focuses on updating and optimizing neural network structures based on the performance of trained networks.

In this method, a computer system with a processor and memory loads a computer-readable representation of a neural network architecture, known as a genome. This genome consists of a set of genes that represent the parameters and properties of the neural network. Using this genome, the computer system generates a neural network with multiple neurons according to the specified architecture.

Once the neural network is generated, it is deployed in a controller for a system or device. The network is then trained by supplying inputs to an input processing unit and receiving outputs from an output processing unit. During training, the synaptic weights of connections between neurons are updated based on the responses to the inputs or neuronal activities within the network.

After training, the performance of the neural network architecture is evaluated. Neuraville’s method involves generating an updated computer-readable representation of an updated neural network architecture based on this evaluation. The updated representation contains an updated genome with modified genes representing updated parameters and properties of the network.

One notable feature of this method is the input processing unit, which generates input spiking signals to the neural network based on a set of input processing unit parameters. These signals are derived from inputs received from an input device, including external stimuli. The input processing unit can be updated with new parameters in the updated architecture.

Neuraville’s approach also allows for the creation or elimination of neurons and synapses during training, offering flexibility and adaptability to the neural network structure.

The computer system maintains a genome repository that stores multiple computer-readable representations of neural network architectures. The updated neural network architecture can be stored in this repository, along with statistical data associated with the performance of the architecture. This repository serves as a valuable resource for future research and development.

By automatically generating and updating neural network architectures based on training and evaluation, Neuraville’s method paves the way for more efficient and effective artificial intelligence systems. The ability to optimize neural network structures in response to performance feedback opens up new possibilities for advancements in various fields, including robotics, machine vision, and more.

Are you developing new technology for an existing application? Did you know your development work could be eligible for the R&D Tax Credit and you can receive up to 14% back on your expenses? Even if your development isn’t successful your work may still qualify for R&D credits (i.e. you don’t need to have a patent to qualify). To find out more, please contact a Swanson Reed R&D Specialist today or check out our free online eligibility test.

Who We Are:

Swanson Reed is one of the U.S.’ largest Specialist R&D tax advisory firms. We manage all facets of the R&D tax credit program, from claim preparation and audit compliance to claim disputes.

Swanson Reed regularly hosts free webinars and provides free IRS CE and CPE credits for CPAs. For more information please visit us at or contact your usual Swanson Reed representative.

Recent Posts